AI Eats Product, CEOs Shift, Social Moats

AI Eats Product, CEOs Shift, Social Moats

The Throughline Brief · April 26th. 6 min 7 sec.

Cold open

AI is eating the product org. Top founders are officially calling the end of traditional product management. AppLovin is now using AI for up to 90 percent of its code. The people at the very top are seeing engineers take total ownership. The old structure is quietly fading away. The execution phase is being automated right in front of us.

Intro

This is The Throughline Brief for April 26th. I went through every single major AI and tech podcast this week so you do not have to. Here is the stuff that actually mattered for founders and operators. I am West. Let us get into it.

AI Eats Product

We need to talk about how AI is literally dissolving the product org. This was everywhere this week. On 20VC, Adam Foroughi dropped a massive stat about AppLovin. They use AI for 80 to 90 percent of their code. He predicts the traditional product organization will completely disappear. Engineers will just take on product ownership. Think about that. Foroughi measures success strictly by value creation and business KPIs. Volume of code is entirely irrelevant now. Then you have Evan Spiegel on Lenny's Podcast. He says AI now lets designers directly ship working code. It removes friction. It automates entire workflows. The barrier between design and a live product is gone. Elad Gil took this even further on The Tim Ferriss Show. He says AI is shifting the entire software business model. We are moving away from selling software licenses. We are moving toward delivering measurable work product. Units of labor. And Sam Altman backed this up perfectly on Cheeky Pint. Investors are completely changing what they look for. They want idea guys. People who deeply understand users and market problems. Because the actual execution part is increasingly automated. Here is why that matters to you as a founder. The premium is shifting entirely to deep market insight. Your true moat relies completely on understanding exactly what problem needs solving. The execution is getting commoditized. You need to focus purely on value creation.

The Brutal Physics of AI

Let us shift to the physical realities of scaling AI. The infrastructure limits are hitting hard. Reiner Pope broke this down beautifully on the Dwarkesh Podcast. He looked at Gemini 3.1 and its 50 percent price jump for longer contexts. That signals massive memory bandwidth bottlenecks in the underlying hardware. Pope also revealed that ignoring batch size for LLM inference can make your per token cost a thousand times worse. You face a severe trade off between latency and cost. Pope also pointed out a crazy hardware constraint. Communication between GPU racks is roughly eight times slower than communication within a single rack. That is a critical bottleneck for Mixture of Experts models. Scaling these things is a brutal physical engineering challenge. Then you have the human infrastructure costs. Elad Gil revealed the extraordinary economic reality of AI talent right now on The Tim Ferriss Show. Top AI researchers are experiencing what he calls a personal IPO. We are talking about compensation packages reaching tens to hundreds of millions of dollars. A single researcher pulling down a massive fortune. This intense bidding war is a completely rational strategy to win the race for AI dominance. The economic importance of this technology justifies those massive numbers. As an operator, you have to understand these constraints. The compute is hitting physical limits, and the top talent is commanding generational wealth. Building in this space requires navigating incredibly steep structural costs.

Scott Galloway

Scott Galloway went deep on the modern isolation crisis while talking with Andrew Huberman on the Huberman Lab. They were unpacking how big tech algorithms actively pull young people away from real world engagement. Galloway offered a highly provocative perspective on individual health versus societal well being. He said, I think personally the risks to a 25 year old liver are dwarfed by the risks of social isolation. That is a heavy statement. It forces a complete reframe of how we view public health priorities. The digital isolation engineered by modern tech platforms creates massive psychological damage. Builders need to recognize this reality and design products that actually bring people together in the physical world.

Adam Foroughi

Adam Foroughi sat down on 20VC to discuss the incredible efficiency gains his team is seeing at AppLovin. He detailed how AI is fundamentally changing the way they build software right now. He made a bold prediction about the future of tech teams. He stated, the traditional product organization will dissolve as engineers take on product ownership. This directly upends decades of conventional tech organizational structure. We are watching the lines between engineering, design, and product management completely blur. As an operator, you must prepare your teams for this exact shift. The most valuable people in your company will be engineers with deep product intuition who can ship end to end alone.

Evan Spiegel

Evan Spiegel shared some hard won lessons from the early days of Snapchat while speaking on Lenny's Podcast. He was reflecting on defensibility and how quickly competitors can clone new features. He admitted something surprisingly candid for a social app founder. He said, 15 years ago, we essentially learned that software is not a moat. This truth is incredibly relevant today in the age of generative AI. When code can be instantly generated, software alone provides zero lasting defense. You have to build moats through deep network effects, unique hardware ecosystems, or highly trusted brands. Founders must look beyond their codebase to secure real, durable market power.

The bottom line

This week, ambitious builders are grappling with massive, foundational shifts. AI is actively redefining roles and collapsing the traditional product org right before our eyes. At the same time, we are hitting the harsh physical constraints of scaling cutting edge tech. And there is an urgent call to counteract big tech's isolating effects by finding true purpose and fostering real human connection. That is your Throughline. See you next Sunday.